The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of instructions designed to bypass safety filters or override agent constraints. The use of 'CRITICAL' and 'IMPORTANT' tags is strictly for valid instructional guidance on data analysis logic.- [DATA_EXFILTRATION]: No network operations or sensitive file path access detected. Data operations are limited to standard pandas CSV reading.- [REMOTE_CODE_EXECUTION]: The skill does not download or execute scripts from remote sources. All code snippets provided are templates for standard statistical analysis.- [OBFUSCATION]: All content is in plain text and standard Python. No Base64, zero-width characters, or homoglyph attacks were found.- [INDIRECT_PROMPT_INJECTION]: The skill processes external data (CSV files), which represents an attack surface; however, the skill lacks dangerous capabilities like network access or shell execution that could be exploited via data injection.- [SAFE]: All identified Python dependencies (statsmodels, scikit-learn, lifelines, pandas, numpy, scipy) are well-known, established libraries in the scientific computing community.

tooluniverse-statistical-modeling